hurricane michael

Hurricane Michael, 17 – 20 October 2000
Part I – Summary Report
and Storm Impact on Canada
Part II – Forecast and Warning Critique
Christopher T. Fogarty
Meteorological Service of Canada, Atlantic Region
Science Report Series 2002-01
June 2002
Newfoundland Weather Centre
Meteorological Service of Canada –
Atlantic Region
Environment Canada
P.O. Box 370
Gander, NF A1V 1W7
ISBN 0-662-32616-4
Catalogue No. En57-36/2002-1E
ii
Table of Contents
Part I – Summary Report and Storm Impact on Canada
ABSTRACT......................................................................................................................................................1
1. Synoptic History ..........................................................................................................................................1
2. Meteorological Statistics............................................................................................................................12
3. Damage and Impacts..................................................................................................................................18
4. Forecast and Warning Summary................................................................................................................19
5. Convair Research Flight ............................................................................................................................20
6. Summary ....................................................................................................................................................21
Acknowledgments...........................................................................................................................................21
References.......................................................................................................................................................22
Part II - Forecast and Warning Critique
ABSTRACT....................................................................................................................................................23
1. National Hurricane Center Tropical Discussions........................................................................................23
2. Canadian Hurricane Centre Prognostic Messages ......................................................................................24
3. Newfoundland Weather Centre Public and Marine Forecasts ....................................................................25
3.1 Public Forecasts....................................................................................................................................26
3.2 Marine Forecasts ..................................................................................................................................26
3.3 Special Weather Statements ..................................................................................................................27
4. Forecast Verification Statistics ...................................................................................................................27
4.1 National Hurricane Center / Canadian Hurricane Centre ...................................................................27
4.2 Newfoundland Weather Centre .............................................................................................................30
5. Coordination between the Hurricane Centres and the Newfoundland Weather Centre..............................34
6. Conclusions................................................................................................................................................34
References.......................................................................................................................................................35
iii
iv
Hurricane Michael, 17 - 20 October 2000
Part I - Summary Report
and Storm Impact on Canada
Christopher T. Fogarty, Newfoundland Weather Centre, Gander, NF
ABSTRACT
On 15 October, 2000, a subtropical depression formed north of the Bahamas as an upperlevel low drifted southeast from the eastern United States. After several days being quasi-stationary
over warm water (> 27°C), the system developed tropical characteristics and eventually became
Hurricane Michael on 17 October. Michael began to track north-northeastward on 18 October
ahead of a mid-tropospheric trough approaching from the northwest. During the morning of 19
October, Michael underwent rapid deepening after moving into the baroclinically-energetic region
of the mid-latitude trough. The supply of this energy was enough to counteract the weakening
which would have most certainly occurred over the colder waters south of Newfoundland.
Consequently, Michael made landfall on the south coast of Newfoundland as a 75-knot category
one hurricane with a central pressure of 965 mb.
This report, which is Part I of a two-part report, presents a comprehensive description of this
unusual event of a hurricane re-intensifying at high latitudes, then making landfall in Canada.
Section 1 outlines the synoptic history of the storm, section 2 contains meteorological data, section
3 summarizes damages and impacts, section 4 presents a forecast and warning summary, and
information on the Convair 580 research flight appears in section 5. The report concludes with a
summary in section 6.
1. Synoptic History
Between 13 and 15 October, 2000, an area of low pressure developed north of the Bahamas
as an upper-level cold low drifted southeast over an old surface frontal zone. Satellite imagery
revealed a wrapped-up cloud pattern around the low that was nearly stationary during that period.
On 15 October the circulation appeared to be getting tighter some 500 km west-southwest of
Bermuda. Late that night, the system was declared to be Tropical Depression 17 by the U.S.
National Hurricane Center (NHC) as convection was beginning to wrap around the low’s core. By
the morning of 17 October the system had reached tropical storm strength with a ragged ring of
convection enclosing an eye-like feature. Consequently, Tropical Depression 17 was renamed
Tropical Storm Michael. The NHC storm track for Michael is shown in Figure 1. A “re-analysis”
track will be presented in section 2 using various data collected from the storm while over Canadian
territory.
Michael reached hurricane strength by late afternoon on 17 October based on satellite
imagery and aircraft data. Convection was becoming more abundant and the upper level outflow
pattern was improving. On 18 October the storm began moving north-northeastward ahead of a
mid-latitude trough to the northwest. Early on 19 October the eye had disappeared on satellite
imagery and it was becoming apparent that the storm was being sheared by the upper-level flow
(see 500 mb flow in Figure 2). However, later that morning, a solid band of convection began
wrapping around the storm (see Figure 3) and the eye subsequently returned. At this point Michael
1
was moving northeastward at approximately 35 knots, and a new low associated with the
negatively-tilted trough over the U.S. Northeast formed about 150 km south of Cape Sable, Nova
Scotia (see Figures 2 and 4). Additional information on this storm is available from Franklin et al.
(2001) and Stewart (2000).
FIGURE 1 - National Hurricane Center track map for Hurricane Michael.
2
FIGURE 2 - GEM 00 HR PROG of 500 mb height and vorticity. Baroclinic low surface
position marked with “L”. Hurricane surface position marked with “M”.
3
FIGURE 3 - GOES visible image of Michael at 1532 UTC 19 October.
4
FIGURE 4 - Map of locations cited in the text.
5
During the afternoon of 19 October, a remarkable interaction occurring between the
energetic trough and Michael resulted in explosive deepening. It is estimated that the deepening
was equivalent to a rate of over 2 mb per hour in the six hours between 1200 UTC and 1800 UTC
19 October. A ship, the MSC Xingang (call sign 3EHR6), located just east of the centre at 1700
UTC reported a sea level pressure of 965.5 mb and a sustained southerly wind of 80 knots
(anemometer height 30 metres - National Weather Service (2002)). At the time, Michael was
racing northeastward at approximately 60 knots indicating just how quickly the cyclone was picked
up by the upper-level flow. As the hurricane tracked toward the south coast of Newfoundland, the
original extratropical low south of Nova Scotia became incorporated into the hurricane’s
circulation. Based on map data at 1800 UTC 19 October (see Figure 5) the low may have
maintained its entity to the south-southwest of Michael’s centre, with Michael located in the warm
sector of the baroclinic wave shown. There was a line of convection and lightning along the warm
front, while no convection was observed along the cold front. A new upper-level front formed in
the warm airmass east of the system and extended well into the subtropics. It is believed that this
was more like a moisture convergence zone acting as a stream for tropical moisture into the midlatitudes (see the satellite image in Figure 6). There was also an abundance of convection with
lightning along this front. This “hybrid” storm reached the south coast of Newfoundland just west
of Harbour Breton around 8:00 p.m. local time (2230 UTC).
FIGURE 5 - Surface re-analysis at 1800 UTC 19 October
6
FIGURE 6 - GOES infrared satellite at 1815 UTC 19 October.
At landfall, radar imagery from Holyrood, Newfoundland (about 250 km to the east)
revealed a continuous arc of echoes encircling over half of the north side of the centre (see Figure
7). This was believed to be the weakening eyewall. Convection was shallow, and almost nonexistent over the southwestern quadrants of the storm. A satellite image (see Figure 8) just two
hours prior to landfall shows the inner core convection beginning to erode from the southern side of
the storm, exposing the low-level centre. As the storm moved inland, all deep-layered cloud and
convection rushed northward leaving the characteristic comma-shaped pattern of an occluded
extratropical cyclone. The trough which had accelerated Michael northward had formed a closed
low over Prince Edward Island by 0000 UTC 20 October (see Figure 9) and was acting as a brake,
causing the now extratropical storm to slow significantly during the six hours after landfall. By
0600 UTC, Michael had moved to lie just north of the Baie Verte peninsula. Surface analyses and
observations show that the central pressure of 965 mb at landfall was maintained through the night
as the storm moved inland. The sea level pressure field had expanded significantly, however,
during that time. It was noted that the diameter of the 980-mb contour doubled between 0000 UTC
and 0600 UTC 20 October (see Figure 10). It was also noted that at 0000 UTC 20 October just
after landfall, the extratropical low which had originally formed off Nova Scotia may have still been
present (see Figure 10a). Data from the Burgeo Bank Buoy (44255) indicated winds shifting to
7
northwesterly as Michael passed, then briefly to southeasterly between 0000 and 0100 UTC 20
October with a second small drop in sea level pressure. A similar pattern was observed three hours
later at Sagona Island (WZN). These observations substantiate the theory that the extratropical low
may have trailed Michael during the transition, and did not become completely absorbed until after
0000 UTC 20 October.
As the remnant hurricane became collocated with the upper-level low early on 20 October, it
began to move northeastward and gradually weakened. Tropical characteristics by this time were
virtually non-existent and the circulation of the storm expanded significantly in size.
FIGURE 7 - 1.5 km CAPPI radar image from Holyrood.
8
FIGURE 8. Advanced very high-resolution radiometer colour composite satellite image of
Michael two hours before estimated time of landfall. Estimated surface position shown with
an asterisk. (Image courtesy of the Ocean Remote Sensing Group, Johns Hopkins University
Applied Physics Laboratory)
9
FIGURE 9 - GEM 00 HR PROG of 500 mb height and vorticity.
extratropical Michael marked with "L".
10
Surface position of
a)
b)
FIGURE 10 - Surface re-analyses at a) 0000 UTC and b) 0600 UTC 20 October.
11
2. Meteorological Statistics
A re-analysed track was determined for Hurricane Michael using various sources. These
include data from the NHC best track archive (see storm track in Figure 1), track positions included
in the Canadian Hurricane Centre (CHC) bulletins and data from all available ships, buoys, and
surface weather stations used to “fine-tune” the track data during the period after 1200 UTC 19
October. Information on sea surface temperatures (SSTs) and forward speed of motion has also
been added to the conventional track information. These data appear in Table 1 and are plotted on
the maps in Figure 11.
TABLE 1 - Track data for Hurricane Michael
LAT
30.00
30.00
29.90
29.90
29.70
29.80
29.90
29.80
29.80
30.10
30.40
30.80
31.50
32.60
34.20
36.30
39.50
41.30
44.00
46.30
47.60
48.50
49.50
50.20
51.00
52.00
LON
-71.20
-71.50
-71.80
-71.90
-71.70
-71.40
-71.10
-71.00
-70.90
-70.90
-70.90
-70.80
-70.40
-69.50
-67.80
-65.50
-62.00
-60.20
-58.90
-57.00
-56.50
-56.50
-56.30
-55.80
-53.50
-50.30
TIME-UTC
10/15/12Z
10/15/18Z
10/16/00Z
10/16/06Z
10/16/12Z
10/16/18Z
10/17/00Z
10/17/06Z
10/17/12Z
10/17/18Z
10/18/00Z
10/18/06Z
10/18/12Z
10/18/18Z
10/19/00Z
10/19/06Z
10/19/12Z
10/19/15Z
10/19/18Z
10/19/21Z
10/19/2230
10/20/00Z
10/20/03Z
10/20/06Z
10/20/12Z
10/20/18Z
WIND-KT
30
30
35
35
35
35
35
45
55
65
65
65
65
70
75
65
75
75
85
80
75
65
65
60
60
60
PRES-MB
1007
1006
1005
1005
1005
1004
1003
1000
995
988
988
986
984
979
983
986
979
975
965
965
965
966
966
966
968
970
STAT
ST DEP
ST DEP
STS
STS
STS
STS
TS
TS
TS
HRCN-1
HRCN-1
HRCN-1
HRCN-1
HRCN-1
HRCN-1
HRCN-1
HRCN-1
HRCN-1
HRCN-2
HRCN-1
HRCN-1
XT-1
XT-1
XT
XT
XT
SST-C SPEED-KT
29
2 ENE
29
3 W
29
3 WSW
29
1 W
28
3 SE
28
3 ENE
28
3 ENE
28
1 SE
28
1 E
28
2 N
28
3 N
27
4 N
27
8 NNE
27
13 NNE
26
21 NE
25
28 NE
25
42 NE
22
45 NE
16
58 NNE
13
53 NNE
LANDFALL
INLD
44 N
INLD
20 N
10
15 NNE
8
17 ENE
8
22 ENE
LEGEND
Storm Status (STAT):
ST DEP – Subtropical Depression
STS – Subtropical Storm
TS – Tropical Storm
HRCN-1 – Hurricane Category 1 on the Saffir-Simpson Scale
HRCN-2 – Hurricane Category 2 on the Saffir-Simpson Scale
XT-1 – Extratropical Storm (Hurricane-force winds)
XT – Extratropical Storm (Storm-force winds)
WIND – Maximum sustained wind speed
PRES – Minimum central pressure
SST – Estimated Sea Surface Temperature beneath the eye/centre
SPEED - Translational speed and heading determined from previous 3- or 6-hr
positions
Intermediate times are boldface
12
FIGURE 11 - Re-analysis track maps for Hurricane Michael for (a) the entire life cycle and (b)
the latter period including re-intensification and transition. Times are UTC [mmddhh].
Michael passed close to Sable Island, ship 3EHR6, buoys 44255 and 44139, and a few
automated surface stations in Newfoundland (see Figure 4). These data proved crucial in the
issuing of warnings, production of surface analyses, and determination of wind speeds around the
hurricane. A very crucial observation was from a ship (call sign 3EHR6) located about 120 km
south-southeast of Sable Island at 1700 UTC 19 October:
Id
TDDHH
3EHR6 1917
Lat
43.0N
Long Wind Cld
59.4W 1880
/
VVWX
0R+
Air Dew Sea Pressure DV
20.0 20.0
655 \550 W4
Until the time of this observation, the estimated minimum pressure was 975 mb with
maximum sustained winds of 75 knots (1500 UTC 19 October). According to satellite imagery, the
ship was located in the eastern portion of the eyewall; therefore, Michael was promptly upgraded to
a category two hurricane with maximum sustained winds of 85 knots. This was the only
observation from this particular ship during the event.
Note that the re-analyzed track takes a slight displacement to the west at the 18 UTC 19
October position, which was also around the time when Michael was meeting up with the baroclinic
low to the northwest (see Figure 5). It is hypothesized that this slight “jog” to the west may have
been in response to the two lows interacting. In the 1800 UTC surface analysis, the extratropical
low may have been located in the trough to the southwest of Michael.
13
The storm tracked roughly midway between the Banquereau Bank buoy (44139—see Table
2) and Sable Island (CWSA—see Table 3) at 1800 UTC 19 October. These data were very useful
in determining the pressure gradient near the centre and the evolution of the sea state. Significant
wave height data from the buoy showed that seas gradually built to 4 m ahead of the storm then
jumped to near 6 m as it passed. In the southwesterlies behind Michael, seas further built to near 8
m. At the Nickerson Bank Buoy (44251) about 30 km southwest of Cape Race, Newfoundland, seas
built to near 5 m around the time Michael made landfall and to 7.5 m in its wake. At this point, the
hurricane was moving too rapidly to allow for substantial wave growth, as noted in the evolution of
seas at these buoys.
TABLE 2 - Observations from Moored NOMAD Buoy 44139, October 19-20
Id
44139
44139
44139
44139
44139
44139
44139
44139
44139
44139
44139
TDDHH
1915
1916
1917
1918
1919
1920
1921
1922
1923
2000
2001
Lat
44.3N
44.3N
44.3N
44.3N
44.3N
44.3N
44.3N
44.3N
44.3N
44.3N
44.3N
Long
57.4W
57.4W
57.4W
57.4W
57.4W
57.4W
57.4W
57.4W
57.4W
57.4W
57.4W
Wind
1317/19
1525/31
1429/36
1442/54
2042/56
2129/34
2238/48
2238/54
2342/56
2438/48
Air
16.0
17.2
17.7
18.8
17.3
16.2
15.3
14.6
14.0
12.7
Dew
13.0
12.9
12.9
12.9
13.1
13.0
12.9
12.8
13.0
13.3
Sea
Pressure DV Wave
980 \081 0 3.7
945 \091 0 3.4
897 \113 0 3.3
835 \146 0 3.9
- 857 \040 0 5.7
882 /046 0 5.1
895
0 6.0
905 /049 0 6.9
920 /037 0 7.7
942 /045 0 7.9
TABLE 3 - Observations from Sable Island (CWSA)
METAR CWSA 191700Z AUTO
HR PRESFR SLP796
SPECI CWSA 191709Z AUTO
SPECI CWSA 191736Z AUTO
SPECI CWSA 191737Z AUTO
SPECI CWSA 191744Z AUTO
SPECI CWSA 191749Z AUTO
SPECI CWSA 191757Z AUTO
SPECI CWSA 191759Z AUTO
METAR CWSA 191800Z AUTO
SLP789
SPECI CWSA 191804Z AUTO
SPECI CWSA 191806Z AUTO
SPECI CWSA 191811Z AUTO
SPECI CWSA 191818Z AUTO
SPECI CWSA 191821Z AUTO
SPECI CWSA 191838Z AUTO
06010KT 3/8SM -RA VV001 17/16 A2892 RMK PCPN
1.5MM PAST
03012KT 1/8SM -RA VV001 17/16 A2889 RMK PRESFR
28019G25KT 1SM OVC001 16/15 A2892
28021KT 5/8SM -RA VV001 16/15 A2893
27028G35KT 3/8SM +RA VV001 17/16 A2896
28021G35KT 1/4SM +RA VV003 17/16 A2896
30018G25KT 3/8SM +RA VV004 16/15 A2892
31018G24KT 3/4SM RA VV004 16/15 A2890
32021KT 1SM RA VV004 16/15 A2890 RMK PCPN 7.0MM PAST HR
32020G26KT
33021G26KT
33017G25KT
27018G25KT
25018G27KT
26015G20KT
1SM -RA
1SM -RA
1 1/8SM
1 3/8SM
1 3/8SM
2 1/2SM
VV002 16/15 A2887
VV001 16/15 A2885 RMK PRESFR
-RA VV001 16/15 A2883 RMK PRESFR
-RA VV001 16/15 A2895
-RA VV002 16/15 A2896
FEW003 FEW021 16/15 A2897 RMK PRESRR
Details of the time, location, and intensity at landfall are seen through an examination of the
data from buoy 44255 (Table 4), the staffed station (LFVP) at St. Pierre (Table 5) and three
automated surface stations (WDS, WBF, WZN - see Table 6). Key observations are boldface.
14
TABLE 4 - Observations from Canadian Moored Buoy 44255 (October 19–20)
TDDHH
44255
44255
44255
44255
44255
44255
44255
44255
44255
44255
Lat
1918
1919
1920
1921
1922
1923
2000
2001
2002
2003
Long
47.3N
47.3N
47.3N
47.3N
47.3N
47.3N
47.3N
47.3N
47.3N
47.3N
57.4W
57.4W
57.4W
57.4W
57.4W
57.4W
57.4W
57.4W
57.4W
57.4W
Wind
0633/42
0633/44
0536/46
0436/48
0031/44
3419/25
3203/09
1303/07
2611/23
2619/23
Air
7.9
8.0
8.1
8.5
10.1
9.8
9.2
8.7
8.3
7.5
Dew Sea
9.4
8.9
8.9
8.9
8.9
9.0
9.0
9.0
9.0
8.9
Pressure
950 \123
902 \133
837 \158
760 \188
677 \225
725 \113
762 /002
765 /088
762 /037
785 /022
TABLE 5 - Observations from Staffed Station St.Pierre (LFVP) on October 19
LFVP 71805
METAR LFVP
METAR LFVP
METAR LFVP
METAR LFVP
METAR LFVP
ST. PIERRE FR 46.76N 56.17W 5m
191800Z 09023KT 1500 RA BR BKN003 09/09 Q0997
191900Z 08024KT 1500 RA BR BKN003 10/09 Q0991
192000Z 08022KT 2000 -RA BR OVC003 10/10 Q0986
192100Z 12031G44KT 2000 -RA BR OVC008 15/14 Q0977
192200Z 18051G63KT 2000 R26/0600N -RA BR BKN006 15/14 Q0974
TABLE 6 - Observations from Automatic Stations WDS, WBF, WZN on October 19–20
WDS 71802 ST. LAWRENCE NF CN 46.92N 55.38W 49m
Following data from Thursday October 19 2000
WDS SA 1800 AUTO8 M M M 995/10/10/1026+43/M/ PK
WDS SA 1900 AUTO8 M M M 944/11/11/0931+50/M/ PK
WDS SA 2000 AUTO8 M M M 891/13/13/1035+44/M/ PK
Following data from Friday October 20 2000
WDS SA 2100 AUTO8 M M M 836/16/16/1132+44/M/ PK
WDS SA 2200 AUTO8 M M M 813/18/08/1752+74/M/ PK
WDS SA 2300 AUTO8 M M M 813/17/01/1944+60/M/ PK
WDS SA 0000 AUTO8 M M M 822/16/01/2045+59/M/ PK
WDS SA 0100 AUTO8 M M M 823/14/04/2142+59/M/ PK
WDS SA 0200 AUTO8 M M M 812/13/05/2032+45/M/=
WDS SA 0300 AUTO8 M M M 806/13/13/2340+68/M/ PK
WDS SA 0400 AUTO8 M M M 828/10/03<2443+57/M/ PK
WDS SA 0500 AUTO8 M M M 845/09/05/2445+68/M/ PK
WDS SA 0600 AUTO8 M M M 856/08/06/2547+65/M/ PK
WND 0953 1736Z=
WND 0850 1850Z=
WND 1044 1952Z=
WND
WND
WND
WND
WND
1045
1574
1793
1972
2259
2009Z=
2157Z=
2210Z=
2302Z=
0057Z=
WND
WND
WND
WND
2268
2360
2474
2571
0255Z=
0316Z=
0446Z=
0532Z=
WBF 71194 BURGEO 2 (AU8) NF CN 47.62N 57.62W 12m
Following data from Thursday October 19 2000
WBF SA 1800 AUTO8 M M M 970/08/07/0833+41/M/0024 PK WND 0944 1716Z=
WBF SA 1900 AUTO8 M M M 913/07/06/0839+49/M/0040 PCPN 4.0MM PAST HR
1848Z=
WBF SA 2000 AUTO8 M M M 866/07/07/0837+47/M/0120 PCPN 8.0MM PAST HR
1906Z=
WBF SA 2100 AUTO8 M M M 803<08/07/0743+56/M/0191<PCPN 7.1MM PAST HR
2034Z /R19/=
WBF SA 2200 AUTO8 M M M 727<08/07/0543+67/M/0259<PCPN 6.8MM PAST HR
2157Z=
WBF SA 2300 AUTO8 M M M 725/08/07/0335+56/M/0297<PCPN 3.8MM PAST HR
2208Z /R30/=
15
PK
WND 0850
PK WND 0955
PK WND 0761
PK WND 0767
PK WND 0566
Following data from
WBF SA 0000 AUTO8 M
WBF SA 0100 AUTO8 M
WBF SA 0200 AUTO8 M
WBF SA 0300 AUTO8 M
WBF SA 0400 AUTO8 M
0354Z=
WBF SA 0500 AUTO8 M
WBF SA 0600 AUTO8 M
0556Z=
Friday October 20 2000
M M 751/09/07/0427+44/M/0297<PK WND 0262 2322Z=
M M 756/08/06/0321+29/M/ PK WND 0347 0003Z=
M M 764/07/07/3503/M/ PK WND 0232 0103Z=
M M 778/07/06/2809/M/ PK WND 2819 0205Z=
M M 800/06/06/2820+29/M/0011 PCPN 1.1MM PAST HR PK WND 2829
M M 822/06/05/2817+29/M/0011 PK WND 2729 0442Z=
M M 838/05/05/2828+40/M/0018 PCPN 0.7MM PAST HR PK WND 2940
WZN 71408 SAGONA ISL (AU4) NF CN 47.37N 55.80W 35m
Following data from Thursday October 19 2000
WZN SA 1847 AUTO4 M M M 967/08/M/0741+48/M/0122 PCPN 3.0MM PAST HR=
WZN SA 1947 AUTO4 M M M 910/08/M/0543+50/M/0118 PCPN 11.8MM PAST HR=
WZN SA 2047 AUTO4 M M M 848<09/M/0548+57/M/0150 PCPN 3.2MM PAST HR=
WZN SA 2147 AUTO4 M M M 776<14/M/1047+63/M/0200 PCPN 5.0MM PAST HR=
WZN SA 2247 AUTO4 M M M 754/14/M/1569+81<M/0204 PCPN 0.4MM PAST HR=
WZN SA 2347 AUTO4 M M M 786/13/M/1861+73<M/0204 PRESRR 5062 ?8MMM?=
Following data from Friday October 20 2000
WZN SA 0047 AUTO4 M M M 792/12/M/1855+65<M/0204 3016 ?6MMM?=
WZN SA 0147 AUTO4 M M M 790/11/M/1850+58/M/ 0036 ?7MMM?=
WZN SA 0247 AUTO4 M M M 779/11/M/1845+52/M/ 8007 ?7MMM?=
WZN SA 0347 AUTO4 M M M 792/08/M/2247+53/M/ 5000 ?1MMM?=
WZN SA 0447 AUTO4 M M M 799/08/M/2252+63<M/0002 PCPN 0.2MM PAST HR=
WZN SA 0547 AUTO4 M M M 810/07/M/2355+63<M/0002 1031 ?8MMM?=
WZN SA 0647 AUTO4 M M M 824/07/M/2347+57/M/0002 3032 ?5MMM?=
Data for Sagona Island (WZN) have been graphed (see Figure 12) to show the trends in sea
level pressure, winds, and temperature as the hurricane passed approximately 80 km to the west. A
sudden increase in wind accompanied by an abrupt rise in temperature occurred at this time. The
pressure dropped rapidly (~ 6mb/hr) up until 2247 UTC, but afterwards only rose gradually. In fact,
the pressure actually dropped slightly to 978 mb near 0300 UTC. This pattern was observed at
other stations as well. As mentioned before, this second minimum was likely associated with the
passage of the low or trough extending to the south-southwest of Michael.
16
Hourly SFC Data - Sagona Island (WZN)
47.37N 55.80W 35m
1005
80
1000
70
995
60
Sea Level Pressure (mb)
990
50
985
40
pressure
980
30
975
Sustained Wind Speed (kts)
Temperature (°C)
wind
20
970
temperature
10
965
20/1447
20/1347
20/1247
20/1147
20/1047
20/0947
20/0847
20/0747
20/0647
20/0547
20/0447
20/0347
20/0247
20/0147
20/0047
19/2347
19/2247
19/2147
19/2047
19/1947
19/1847
19/1747
0
19/1647
960
Tim e UTC (DD/HHMM)
FIGURE 12 - Time series of surface data from Sagona Island (WZN)
The estimated time of landfall was 2230 UTC 19 October at 47.6N 56.5W (about 60 km
west of Harbour Breton, Newfoundland). The strongest sustained winds reported by a land station
were 69 knots with gusts to 81 knots from Sagona Island at 2247 UTC. Note that this station only
reported data once every hour. The sustained wind is a 2-minute mean and the gusts represent the
highest wind in the previous 10 minutes. A peak wind gust of 93 knots from St. Lawrence (WDS)
at 2210 UTC was the highest wind speed recorded. The lowest pressure observed near landfall was
967.7 mb from the Burgeo Bank buoy (44255) at 2200 UTC with an incredible 22.5-mb drop in the
preceding three hours. The estimated minimum pressure at landfall was 965 mb.
Rainfall amounts and rates from Michael were not quite typical of those found in category
one hurricanes. At the time of landfall, Michael was moving very fast (approximately 50 kts) and
convection—although still somewhat organized in the remnant eyewall—was becoming shallower.
This is no surprise since SSTs near the coast were only 13°C (55°F). The rule of thumb used by
forecasters at the National Hurricane Center for estimating rainfall amounts in hurricanes is to
divide 2500 by the forward speed of the storm in knots. In Michael’s case, 2500 / 50 kt = 50 mm.
Amounts near 30 mm (a little over one inch) fell along the track while amounts up to 75 mm (3
inches) were recorded to the west of the track over western and northern Newfoundland. The
higher amounts in these locations were attributed to the large-scale forcing associated with the
interacting trough over the Gulf of St. Lawrence as well as enhancement owing to elevated terrain.
A summary of total rainfall and maximum wind statistics appears in Table 7.
17
TABLE 7 - Wind and rainfall reports during Hurricane Michael
STATION
SABLE ISL
SAGONA ISL
ST LAWRENCE
ST PIERRE
MARTICOT ISL
WINTERLAND
BURGEO
CAPE RACE
BONAVISTA
GANDER
BADGER
ST JOHN’S
GRATES COVE
TWILLINGATE
PORT AUX BASQUES
STEPHENVILLE
DEER LAKE
LA SCIE
ENGLEE
ST ANTHONY
STATION
WSA
WZN
WDS
LFVP
WTM
XWT
WBF
WRA
WVA
YQX
WDI
YYT
WVW
WDO
WZB
YJT
YDF
WAG
WDA
WDW
ID
WSA
WZN
WDS
LFVP
WTM
XWT
WBF
WRA
WVA
YQX
WDI
YYT
WVW
WDO
WZB
YJT
YDF
WAG
WDA
WDW
MIN PRES
976.2mb
975.4mb
980.6mb
974.0mb
978.8mb
977.4mb
972.5mb
988.4mb
978.3mb
973.6mb
972.0mb
984.8mb
982.6mb
969.8mb
979.4mb
980.0mb
977.6mb
968.2mb
972.0mb
LOCATION
43.93N 60.02W
47.37N 55.80W
46.92N 55.38W
46.77N 56.17W
47.33N 54.58W
48.13N 55.33W
47.62N 57.62W
46.65N 53.07W
48.67N 53.12W
48.95N 54.57W
48.97N 56.07W
47.62N 52.75W
48.17N 52.93W
49.68N 54.80W
47.57N 59.15W
48.55N 58.55W
49.22N 57.38W
49.92N 55.67W
50.72N 56.12W
51.38N 56.10W
TIME(UTC)
19/1811
19/2247
20/0300
19/2200
20/0334
20/0300
19/2300
20/0400
20/0700
20/0500
20/0100
20/0500
20/0500
20/1100
19/2200
20/0300
20/0400
20/1000
20/0800
ELEV
4m
35m
49m
5m
27m
29m
12m
28m
30m
151m
105m
140m
18m
92m
38m
26m
22m
194m
29m
29m
MAX WND/PK GUST
280/36G50KT
150/69G81KT
170/52G93KT
180/51G63KT
230/54G69KT
190/40G61KT
050/43G67KT
260/44G66KT
220/44G73KT
180/32G45KT
270/19G41KT
230/35G63KT
230/46G69KT
270/45G61KT
300/36G46KT
300/28G44KT
060/20G35KT
310/30G55KT
010/36G59KT
PRECIP
30mm
33mm
27mm
34mm
19mm
21mm
19mm
27mm
27mm
17mm
58mm
45mm
45mm
49mm
75mm
-
TIME PK GUST
19/2200
19/2247
19/2210
19/2200
20/0515
19/2306
19/2157
20/1026
20/0700
20/0300
20/1031
20/0940
20/0622
20/1200
20/0600
20/1200
19/2300
20/1305
20/1000
3. Damage and Impacts
For the most part, damage associated with Michael was limited and caused mainly by high
winds and not to coastal or inland flooding. No part of the island was left unaffected, although most
of the damage was confined to a few small communities east of where Michael had made landfall
(see Figure 4). Power outages, lasting a few hours, were reported in Harbour Breton, Gaultois,
Hermitage, and English Harbour West. Many homeowners reported vinyl siding being peeled off,
windows being smashed, and shingles torn from roofs. The community of Gaultois, located about
50 km east of landfall, appeared to sustain the most damage. A single-storey home lost its entire
18
roof. A large segment of the roof at a nearby hotel was also lost. Throughout the town, service lines
were left dangling, trees uprooted, and fences damaged. Senior residents described the wind as
being horrifying and the worst they could remember. Power interruptions were witnessed as far
east as St. John’s and as far west as Corner Brook, where winds were strong enough to break large
tree branches.
On the water, a tug (the Atlantic Elm) in Fortune Bay lost its tow to an unmanned barge (the
Portland Star) around 7:00 p.m. NDT when seas became too rough. The barge, containing 8000
metric tonnes of cement and 10000 litres of diesel fuel, was later reported to have sunk somewhere
in Fortune Bay. Pleasure craft moored at ports around Fortune Bay were damaged or capsized. The
storm surge from Michael caused little coastal flooding since it arrived at low tide. The tide gauge
at Argentia, about 200 km east of landfall, reported a surge near 80 cm, or just less than 3 feet. It is
possible that the surge in Fortune Bay may have been up to 5 feet, but there were no confirming
reports. Passenger ferry service between Newfoundland and Nova Scotia was disrupted. Vessels
had difficulty docking and passengers had to ride out the storm just offshore.
4. Forecast and Warning Summary
Hurricane Michael presented a great challenge for forecasters. The most challenging aspect
of this event was determining how the hurricane would interact with the baroclinic low off Nova
Scotia. The main question was which system would become the dominant feature. Large-scale
numerical model performances on 18 October presented variations on the solution: The Canadian
Global Environmental Multiscale (GEM) model appeared to be developing the remnant system, but
at the expense of Michael. The U.S. Nested Grid Model (NGM) had a similar idea, but less intense
with the baroclinic development. The U.S. Medium Range Forecast (MRF) model maintained
Michael out to 0000 UTC 20 October but had the tropical system well south of the GEM position.
Tropical models used in Miami insisted that increasing shear would cause Michael to weaken.
It was not until the morning of 19 October that there were firm indications that Michael
would end up being the dominant centre. Wind warnings (for gusts to 110 km/h) were issued by the
Newfoundland Weather Centre (NWC) for the Avalon Peninsula, Burin Peninsula, East Coast, and
Central Newfoundland forecast regions. Heavy rainfall advisories (up to 40 mm) were also issued
for the East Coast, Central Newfoundland, South Coast (except the Burin peninsula), and the West
Coast forecast regions. After confirmation with the Canadian Hurricane Centre in Dartmouth on
the validity of ship 3EHR6, the NWC amended the public forecast at 5:30 p.m. local time (2000
UTC). The amendment was to upgrade the wind warning for the South Coast (east of Burgeo),
including the Burin Peninsula, for southerly winds of 70 km/h gusting to 130 km/h for Thursday (19
October) evening. Rainfall amounts were left unchanged. In the 10:30 p.m. (0100 UTC 20
October) public forecast update, Michael was well inland and rainfall advisories were continued
only for the south coast (Burgeo and west), and the West Coast. Wind warnings remained in effect
for westerlies gusting to 110 km/h on the south side of the remnant storm on 21 October.
Storm force wind warnings were in effect for the majority of Newfoundland waters in the
marine forecast issued at 1800 UTC 19 October. Generally, these warnings were for winds of up to
50 knots: first in the southeasterlies ahead of, then in the westerlies behind Michael. As with the
public forecast, a revised marine forecast was sent out at 5:00 p.m. after verification of the
hurricane’s intensity. The amendment was to upgrade the storm warning to a hurricane-force wind
warning (65 to 75 knots) for the Southwest Coast east of Ramea. For the South Coast region,
storm-force winds of 50 knots were upgraded to 60 knots.
In addition to these regular products issued by the NWC, three Special Weather Statements
were issued. The first of these was issued at 9:50 p.m. (0020 UTC 20 October) stating where
19
Michael had made landfall, the maximum observed winds, and the extent of rainfall. Another was
sent out at 12:55 a.m.—essentially, an update of the previous issue—with a new maximum wind
report of 93 knots (172 km/h) at St. Lawrence. A third and final report was sent out at 12:14 p.m.
(1444 UTC 20 October) summarizing wind and rainfall amounts observed across Newfoundland. A
more complete collection of warnings and forecast verifications appears in Part II.
5. Convair Research Flight
As Michael was forming well south of Canadian waters, arrangements were being made by
the Meteorological Research Branch of Environment Canada to send a research plane into Michael
as it was undergoing transition. At 1521 UTC 19 October the “Convair 580”—an aircraft owned
and operated by the National Research Council—departed from Greenwood, Nova Scotia, to
examine the storm (Wolde et al., 2001). The aircraft was equipped with an 8-mm (K-band) radar
and sophisticated sensors used to measure the microphysical properties of clouds. Sixteen GPS
(Global Positioning System) dropsondes were deployed during the 3-hour, 50-minute journey.
Figure 13 shows the sequence of dropsondes that were released while the Convair flew just south of
the storm centre at approximately 6400 m (21,000 feet). Radar data revealed interesting vertical
structure to the clouds that were topped near 10700 m (35,000 feet). Very strong headwinds of 120
knots were experienced on the return leg of the mission. Dropsonde data near the storm centre
revealed a relatively deep layer of 130-knot southwesterly winds from the top of the boundary layer
up to approximately 750 mb (2500 m). Details of this flight, including an in depth analysis of
flight-level and dropsonde data, are planned for an upcoming article in the Bulletin of the American
Meteorological Society. An article summarizing the mission appeared in the winter 2000 issue of
Environment
Canada’s
newsletter
Zephyr
(http://www.mscsmc.ec.gc.ca/cd/zephyr/winter2000/page 20_e.html).
FIGURE 13. Convair dropsonde sequence and Hurricane Michael storm track segment.
20
6. Summary
Hurricane Michael was the third hurricane to make landfall in eastern Canada during the 5year period from 1995 to 2000. Hurricane Luis came ashore on the Avalon Peninsula of
Newfoundland on 11 September 1995 while Nova Scotia experienced the brunt of Hurricane
Hortense on 14 September 1996. Michael was one of four tropical systems to affect Newfoundland
and adjacent marine areas in 2000 (average is one or two per year). On 17 September Tropical
Storm Florence passed just east of the Avalon Peninsula, bringing heavy rainfalls to eastern parts of
Newfoundland. On 25 September the remains of Hurricane Helene crossed the northern Grand
Banks, and on 8 October the remnants of Tropical Storm Leslie crossed central Newfoundland with
heavy downpours between St. John's and Gander.
The highlight of the season, Hurricane Michael, made landfall in Newfoundland at 8:00 p.m.
local time on 19 October just west of Harbour Breton, with maximum sustained winds of 75 knots
(140 km/h) and gusts over 90 knots (165 km/h). Michael also brought between 25 and 75 mm of
rain to the province. As the storm moved northeast of the island on 20 October, hurricane-force
westerly winds (65 knots or 120 km/h) continued to impact some coastal areas.
Evidence that Michael was still a hurricane at landfall include the following observations:
• An apparent eyewall was seen in radar imagery (Figure 7) - note that the perceived centre was
significantly displaced to the northeast of the estimated surface centre but hurricane-force winds
were reaching the surface.
• The inner core appeared to remain very tight as shown in Figure 10a.
• An extratropical low which interacted with Michael may have been present near landfall
suggesting Michael was still a separate entity in the warm airmass (see Figure 10a).
In general, light to moderate damage was reported to homes and trees on the south coast.
Siding and shingles were removed from residential buildings in many cases. There was a report of
a roof being blown off a home in Gaultois and a portion of the roof from a local hotel. Several
small boats were also capsized in Fortune Bay. Newfoundland is no stranger to high winds; homes
are built to withstand winter storms which often produce hurricane-force winds. Minimal amounts
of tree or structural damage were reported in Newfoundland. Had Michael struck Nova Scotia,
damage could have been considerably greater due to increased vulnerability to those extremes in
wind. Damage due to the hurricane-force wind alone may have caused more widespread tree and
property damage resulting in significant power and service disruptions comparable to that
experienced along the eastern seaboard of the United States. It is a scenario like this that justifies
ongoing research into the current and future frequency of hurricanes threatening eastern Canada.
Acknowledgments
The author would like to thank George Parkes, John Parker and Jim Abraham for their
insightful reviews of this article. The author would also like to thank Peter Bowyer for his guidance
in preparing this report and input throughout the data analysis period.
21
References
Franklin, James L., Lixion A. Avila, Jack L. Beven, Miles B. Lawrence, Richard J. Pasch, Stacy R.
Stewart, 2001: Atlantic Hurricane Season of 2000. Monthly Weather Review: Vol. 129, No.
12, pp. 3037–3056.
National Weather Service, 2002: National Weather Service Voluntary Observing Ship Program
35 pp. [Available online at http://titan.nih.gov/utpublic/[email protected].]
Stewart, S. R., 2000: Tropical Cyclone Report Hurricane Michael 17 - 19 October 2000. 13 pp.
[Available online at http://www.nhc.noaa.gov/2000michael.html.]
Wolde, M., D. Marcotte, J. Jordan, J. Aitken, J. Abraham, and J. W. Strapp, 2001: The First
Canadian Experience with Research Flight Operations in Hurricane Extratropical
Transition. Canadian Aeronautics and Space Journal, Vol. 47, No. 3, 179-189.
22
Hurricane Michael, 17 – 20 October 2000
Part II - Forecast and Warning Critique
Christopher T. Fogarty, Newfoundland Weather Centre, Gander, NF
ABSTRACT
This report is a follow-up to Part I “Summary Report and Storm Impact on Canada”. It
summarizes the various forecasts and warnings issued by the four weather offices involved in
monitoring Hurricane Michael: The National Hurricane Center (NHC), Miami, FL; the Canadian
Hurricane Centre (CHC), Dartmouth, NS; the Newfoundland Weather Centre (NWC), Gander, NF;
and the Maritimes Weather Centre (MWC), Dartmouth, NS. Forecast verification statistics for both
track and intensity are outlined for the hurricane offices. Verification of public and marine
warnings issued by the NWC for the island of Newfoundland and adjacent waters is also presented.
Warnings issued by the MWC are also summarized. The report concludes with a summary of
coordination efforts between each of the forecast offices and a conclusion outlining the key
challenges of forecasting this event.
1. National Hurricane Center Tropical Discussions
Issue times WTNT41 (intermediate bulletins italicized):
17/0300Z, 17/0900Z, 17/1500Z, 17/2100Z, 18/0300Z, 18/0900Z, 18/1500Z, 18/2100Z,
19/0300Z, 19/0900Z, 19/1500Z, 19/1800Z, 19/2100Z, 20/0030Z
The NHC began issuing tropical cyclone discussions on Tropical Depression 17 at 0300
UTC 17 October. This depression was estimated to be near tropical storm strength at the time,
based on the fact that there was one rain-free vector of 35 knots reported by a QuikSCAT pass.
Note that re-analysis data (see Table 1, Part I) show the system at tropical storm strength at this
time. The depression was embedded in an upper level low, and it was thought that this would inhibit
any significant strengthening in the short term. In the longer term, a trough was forecast to impart
increasing shear as it approached from the northwest.
In their second advisory at 0900 UTC 17 October it was noted that the cloud pattern of now
Tropical Storm Michael was becoming better organized but convection lacked vertical extent. Such
a pattern would be indicative of a hurricane in the deep tropics. Michael was still not expected to
gather much strength before entering the shearing environment that was expected later in the period.
The following discussions and advisories brought Michael to hurricane intensity mainly in
response to the exceptional organization of the system with a well-defined eye. After receiving
aircraft data from the storm on 17 October, Michael was upgraded to hurricane status in the 2100
UTC discussion. In this issue it was suggested for the first time that Michael might become
absorbed by an extratropical low associated with the mid-tropospheric trough over the U.S.
Northeast. At 0900 UTC 18 October the NHC was predicting that the anticipated merger between
Michael and this baroclinic low would be a shear-dominated process and that the resulting cyclone
would not be significant:
“...WHILE SOME STORMS THAT INTERACT WITH THE WESTERLIES STRENGTHEN AS THEY
ACCELERATE...THE GUIDANCE SUGGESTS THESE WESTERLIES SHOULD PRODUCE TOO MUCH
SHEAR TO ALLOW MICHAEL TO DO THAT...”
23
One day later, in their 0900 UTC 19 October issue, the first indication was made that the
trough interaction would at least maintain a strong storm heading toward Newfoundland, but
forecasters were still unclear as to which system would survive — Michael or the new baroclinic
low off Nova Scotia. A significant adjustment in the intensity of Michael was made in an
intermediate bulletin following the report from ship 3EHR6 and it was clear that Michael was going
to remain the dominant cyclone. An excerpt from their 2100 UTC 19 October discussion follows:
“MICHAEL CONTINUES AS A HURRICANE MAINTAINING ITS COMPACT WARM CORE AND IS
RACING TOWARD THE NORTH-NORTHEAST ABOUT 45 KNOTS AHEAD OF A SHORT WAVE.
THE
HURRICANE IS EXPECTED TO CROSS NEWFOUNDLAND WITHIN THE NEXT 6 TO 12 HOURS AS IT
TRANSITIONS TO A POWERFUL EXTRATROPICAL CYCLONE.
THE SKILL IN FORECASTING THE BEGINNING AND THE END OF THE TRANSITION PROCESS IS
VERY LOW. NEVERTHELESS...EITHER TROPICAL OR EXTRATROPICAL ...HURRICANE FORCE
WINDS AND HEAVY RAINS WILL OCCUR TONIGHT OVER PORTIONS OF NEWFOUNDLAND.
SEE
LOCAL FORECASTS PROVIDED BY ENVIRONMENT CANADA.”
In their final bulletin at 0030 UTC 20 October just after landfall, it was indicated that Michael had
acquired frontal structures just before landfall.
The NHC’s nowcasting guidance provided only a short lead-time for the hurricane-force
winds reaching land due to the challenges they had faced with this storm’s track and reintensification. Their earlier assumption had been that the baroclinic trough would produce too
much shear for there to be any re-intensification. Also, they did not originally anticipate that the
system would reach hurricane status, demonstrating the low skill of numerical models in forecasting
pre- and post-transition intensity change.
2. Canadian Hurricane Centre Prognostic Messages
Issue times FXCN05 (intermediate bulletin italicized):
17/0900Z, 17/1800Z, 18/0000Z, 18/0600Z, 18/1200Z, 18/1800Z, 19/0000Z, 19/0600Z,
19/1200Z, 19/1800Z, 19/2100Z, 20/0000Z, 20/0600Z, 20/1200Z
The CHC began issuing prognostic messages on Michael at 0900 UTC 17 October. This
was a short message and served essentially as an introduction to the storm which was expected to
enter the Canadian response zone. In their second discussion at 1800 UTC 17 October the CHC
was already providing details as to the possible outcome of the expected hurricane-trough merger
later in the forecast period. The Global Environmental Multiscale (GEM) model was indicating that
a baroclinic development would occur in about 48 hours, south of Nova Scotia. Several relevant
points were made at this point: 1) the transition would occur faster than indicated by the American
models; 2) a cut-off low would develop aloft; 3) a large baroclinic system would result, bringing
heaviest rains to the Maritime Provinces and western Newfoundland as it decelerated. In the 0600
UTC 18 October discussion, the CHC still had a firm handle on the forecast which would end up
verifying rather well.
“...BY THURSDAY NIGHT...MAJOR COMBINED BAROCLINIC LOW MOVING MUCH SLOWER OVER
NEWFOUNDLAND...”
Large-scale model performances on 18 October presented variations on the solution. The
GEM model appeared to be developing the remnant system, but at the expense of Michael. The
Nested Grid Model (NGM) had a similar idea, but was less intense with the baroclinic development.
24
The Medium Range Forecast (MRF) model maintained Michael out to 0000 UTC 20 October but
had the tropical system well south of the GEM position. Tropical models used in Miami insisted
that increasing shear would weaken Michael.
Late on 18 October, the CHC continued with the philosophy of a faster track as suggested
by the upper flow in the GEM. There was strong indication that the combination of the baroclinic
low off Nova Scotia and the hurricane (whose vortex was poorly resolved by the GEM) would
result in a system with very high thicknesses over its centre (about 570 dam between 1000 and 500
mb). Since the GEM was putting emphasis on the baroclinic low, the track was too far west — just
brushing Cape Breton, Nova Scotia.
A change in thinking was made by the CHC at 0600 UTC 19 October:
“...GEM SHOWS 572 THKNS OVER AVALON PEN AT 20/00Z...WHICH MAY BE CLOSE TO THE
REMNANTS OF MICHAEL. IT MAY MAINTAIN AN IDENTITY EVEN BEYOND AND NOT BECOME
FULLY MERGED WITH THE DEVELOPING LOW UNTIL NEAR 21/12Z...”
The track was adjusted much further east at this point which would bring hurricane Michael into
Placentia Bay, Newfoundland by 0000 UTC 20 October. The delay in the merger to 1200 UTC 21
October was some 36 hours later than the previous thinking, although it is speculated that this was
an error.
It was not until the morning of 19 October that there were indications Michael would end up
being the dominant centre. The track was adjusted back to the west in the 1800 UTC bulletin. Data
from ship 3EHR6 at 1700 UTC had not yet been confirmed so the hurricane’s intensity was kept at
65 knots in that issue. After discussing this report with the NHC, the CHC upgraded the storm to
80 knots in their 2100 UTC discussion, and a final adjustment to the track was made, with
anticipated landfall west of the Burin Peninsula as a hurricane within a couple of hours. By 0000
UTC 20 October Michael was inland over south-central Newfoundland. CHC declared it posttropical at that point with maximum sustained winds of 70 knots. The first indications of forecast
rainfall amounts (near 60 mm) were made in the 0000 UTC issue. This ended up being
representative for areas west of the track, where between 40 and 75 mm were recorded.
The final discussion was issued at 1200 UTC 20 October when the storm was moving
slowly northeastward off the northeast coast of Newfoundland.
3. Newfoundland Weather Centre Public and Marine Forecasts
Issue sequence of key bulletins:
FPCN22:
19/1757Z
FPCN20:
19/1800Z
FPCN11:
19/1829Z
FPCN20:
19/1939Z
FPCN11:
19/2002Z
FPCN20:
19/2230Z
FPCN22:
19/2250Z
WOCN10:
20/0019Z
FPCN11:
20/0100Z
WOCN10:
20/0325Z
WOCN10:
20/1444Z
Marine Synopsis
Marine Forecast
Public Forecast
Marine Forecast (Amendment)
Public Forecast (Amendment)
Marine Forecast
Marine Synopsis
Special Weather Statement
Public Forecast
Special Weather Statement
Storm Summary
25
This section will focus on warnings issued by the NWC. Although there were also warnings
issued by the Maritimes Weather Centre (MWC) during this event, they will not be discussed in
detail here.
Two important forecasts were issued in the afternoon of 19 October. The marine forecast
(FPCN20 CYQX) was issued at 3:30 p.m. local time (1800 UTC) and the public forecast (FPCN11
CYQX) was issued at 4:00 p.m. (1830 UTC). These forecasts followed the theme that Michael
would weaken to tropical storm strength while passing over colder waters and evolve into a strong
extratropical cyclone. The Hurricane Centres had not yet finished assessing the validity of the ship
observation of 80-knot sustained winds at 1700 UTC.
3.1 Public Forecasts
Wind warnings (gusts to 110 km/h) were in effect for the Avalon Peninsula, Burin
Peninsula, East Coast and Central Newfoundland forecast regions. Heavy rainfall advisories (up to
40 mm) were in effect for the East Coast, Central Newfoundland, South Coast (except the Burin
Peninsula), and the West Coast forecast regions. After confirmation with the NHC on the validity
of ship 3EHR6, the CHC relayed the information to the NWC and an amended public forecast was
sent out at 5:30 p.m. local time (2000 UTC):
“...HURRICANE MICHAEL APPROACHING NEWFOUNDLAND FROM THE SOUTH IS NOT WEAKENING
AS FAST AS EXPECTED. AS A RESULT POSSIBLE DAMAGING WINDS WITH GUSTS TO 130 KM/H
ARE EXPECTED THIS EVENING OVER PARTS OF THE SOUTH COAST. AS WELL HEAVY RAIN IS
EXPECTED THIS EVENING...”
The amendment was to upgrade the wind warning for the south coast, east of Burgeo including the
Burin peninsula for southerly winds of 70 km/h gusting to 130 km/h for Thursday (19 October)
evening. Rainfall amounts were not changed. In the 10:30 p.m. (0100 UTC 20 October) public
forecast update, Michael was well inland and rainfall advisories were continued only for the south
coast Burgeo and west, and the West Coast. Wind warnings remained in effect for westerlies
gusting to 110 km/h on the south side of the remnant storm on Friday.
Other public warnings
Heavy rainfall warnings were issued by the MWC in Dartmouth for all of Nova Scotia and
Prince Edward Island and the Magdelan Islands during this event. Wind warnings were issued for
all of Cape Breton, Kings and Queens counties on Prince Edward Island, and the Magdelan Islands.
No warnings were issued by the New Brunswick Weather Centre. Heavy rainfall (50 mm or more)
fell over all but western Nova Scotia. These amounts were also recorded across Prince Edward
Island, the Magdelan Islands and southeastern New Brunswick. High winds (gusts to 90 km/h or
more) were experienced over the Cape Breton highlands and eastern Prince Edward Island. The
weather conditions experienced throughout these regions were not directly related to Michael but
due to the larger scale circulation encompassing Michael.
3.2 Marine Forecasts
Storm-force wind warnings were in effect for the majority of Newfoundland marine regions
in the marine forecast issued at 1800 UTC 19 October. Generally these were for marginal storm
force-winds of 50 kts - first in the southeasterlies ahead of Michael, then in the westerlies behind
Michael. As with the public forecast, a revised marine forecast was sent out at 5:00 p.m. after
verification of the hurricane’s upgraded intensity. The amendment was to upgrade the Storm
26
Warning to a Hurricane-force Wind Warning (65 to 75 knots) for the Southwest Coast east of
Ramea. For the South Coast region, Storms of 50 knots were upgraded to Storms of 60 knots.
The marine synopsis (FPCN22 CYQX) issued at 3:30 p.m. (1800 UTC) discussed the
position and motion of pressure systems throughout the district, as well as a paragraph on hurricane
information and a summary of warnings. This issue opened with saying “...a ridge of high
pressure...” followed by the hurricane information in the last paragraph before the warning
summary. This style was carried over from previous bulletins. To better emphasize the main event,
it would have been best to open the synopsis with information on the hurricane, then discuss other
features such as ridges and minor troughs (discussion based on level of importance). The synopsis
was not updated with the regular forecast amendment, but given the severity and rarity of the event
and the fact that Storm Warnings were upgraded to hurricane-force winds, updating the synopsis
would have increased the warnings’ effectiveness. The 8 p.m. synopsis was re-written such that the
hurricane information appeared in the first paragraph, and provided a better portrayal of the severity
of the storm, even though by that time the centre was inland. (The NWC also made mention of
higher than normal water levels in their marine synopses from information derived at the CHC.)
Other marine warnings
A variety of marine wind warnings were issued by the MWC for their southeastern forecast
waters including hurricane-force winds for Banquereau Bank, Laurentian Fan and the southeastern
half of the East Scotian Slope at 1414 UTC 19 October. For the East Scotian Slope, this provided a
lead-time of only a couple hours. For Laurentian Fan, the lead-time was about four hours and for
Banquereau, about six hours.
3.3 Special Weather Statements
In addition to the regular bulletins issued by the NWC, three Special Weather Statements
(WOCN10 CYQX header) were issued. The first of these was issued at 9:50 p.m. (0020 UTC 20
October) stating where Michael had made landfall, maximum observed winds and the extent of
rainfall. Another one was sent out at 12:55 a.m. which was essentially an update to the previous
issue, with a new maximum wind report of 93 knots (172 km/h) at St. Lawrence. A third and final
report was sent out at 12:14 p.m. (1444 UTC 20 October) summarizing winds and rainfall amounts
observed across Newfoundland.
4. Forecast Verification Statistics
4.1 National Hurricane Center / Canadian Hurricane Centre
Table 1 summarizes the track and absolute wind error statistics for the Hurricane Centres.
Figure 1 presents these data in graphical format.
27
TABLE 1. Hurricane Michael track and absolute wind error statistics for the NHC and CHC
(Track errors are in nautical miles (NM) and wind speed errors are in knots (KT))
NHC - AVERAGE TRACK FORECAST ERRORS (NM)
FCST HR
ERROR
#CASES
12
63
11
24
133
11
36
202
8
48
247
7
72
299
4
NHC - AVERAGE ABSOLUTE WIND SPEED FORECAST ERROR (KT)
FCST HR
12
24
36
48
72
ERROR
11
14
21
25
25
#CASES
11
11
8
7
4
CHC - AVERAGE TRACK FORECAST ERRORS (NM)
FCST HR
12
24
36
48
ERROR
74
122
139
165
#CASES
12
9
9
6
72
170
3
CHC - AVERAGE ABSOLUTE WIND SPEED FORECAST ERROR (KT)
FCST HR
12
24
36
48
72
ERROR
9
13
15
17
12
#CASES
12
9
9
6
3
Significant challenges with predicting Michael were related to intensity changes both
initially and during the baroclinic transition. The storm’s intensity was under-forecast by both
Centres. Both track and wind speed errors were reported to have been greater than the 10-year
average (1990-1999) at the NHC (Stewart 2000, section C) . Forecast track errors were large
during the pre-transition phase because it was unclear which low centre would dominate - Michael
or the baroclinic low off Nova Scotia.
The CHC was successful in adapting data from the NHC to provide a more accurate forecast
of both the track and wind speeds, especially beyond 36 hours. In addition, the CHC was talking
about the possibility of a re-intensification more than 24 hours sooner than the NHC. This earlier
“jump” on the transition was reflected in the forecast accuracy.
28
Average Track Forecast Errors (nm)
Hurricane Michael
300
250
NHC
Error (nm)
200
CHC
150
100
50
0
12
24
36
48
72
Fcs t Hr
Average Absolute Wind Speed Forecast Error (kt)
Hurricane Michael
30
25
NHC
error (kt)
20
CHC
15
10
5
0
12
24
36
48
72
Fcst Hr
FIGURE 1. Track and absolute wind error statistics for the NHC and CHC
29
4.2 Newfoundland Weather Centre
Table 2 summarizes the warnings and advisories issued by the NWC and their verifications
(underestimated forecast values indicated with ‘U’):
TABLE 2. NWC forecast verification summary
Warning/Advisory Region
Lead Time
Verification
_____________________________________________________________________
inland regions...
Wind Warning
Avalon, Burin
15 hrs
Hit (U)
(60G110 km/h)
East Coast, Central NF
Rainfall Advisory
(30-40 mm)
South Coast (except Burin)
West/East Coast, Central NF
12 hrs
Hit (U)
coastal marine regions...
Gale Warning
Northeast Gulf, Gulf PortAu Port
42 hrs
Hit (U)
Storm Warning
Belle Isle
24 hrs
Hit
Storm Warning
Southwest Coast, South
Coast
39 hrs
Hit (U)
Storm Warning
East Coast, Northeast
Coast
30 hrs
Hit (U)
Hurricane Warning
Southwest Coast (East of Ramea)
2 hrs
Hit
Missed events:
Wind:
South Coast (west of Burin)
Northern Peninsula
Heavy Rainfall:
West Coast, East Coast (Baie Verte), Northern Peninsula (east)
Storm:
Northeast Gulf
The verifying data for rainfall and winds are shown in Figures 2 and 3, respectively. Rainfall
data (Richards, 2000) comprising Figure 2 were collected from a network of 64 rain gauges across
the island. A complete listing of those data appears in Table 3.
30
TABLE 3. Comprehensive post-landfall rainfall data - Hurricane Michael - 06 UTC 19
October - 06 UTC 21 October 2000 for Newfoundland (island)
No. of Sites
64
LAT
LON
DD MM DD MM
PR mm SITE
48
47
49
48
47
48
48
48
47
47
46
49
48
48
49
48
49
50
49
50
48
49
48
48
47
48
47
48
47
50
50
47
47
49
47
48
47
48
51
49
49
49
47
51
49
47
48
47
49
47
49
47
47
47
27.8
40.6
28.8
50.9
35.8
20.5
16.9
33.8
33.9
33.0
19.1
20.8
12.8
13.2
52.4
43.1
15.2
48.0
48.8
77.0
47.0
35.2
60.0
27.8
23.9
28.8
24.0
20.0
36.0
65.4
11.2
12.0
21.2
22.6
43.2
16.0
14.0
39.0
56.0
62.8
76.0
27.8
38.0
41.2
29.2
58.5
20.0
23.0
29.4
41.0
53.5
33.0
28.2
28.0
58
59
02
34
26
40
02
50
37
35
39
24
23
12
19
56
54
14
10
43
46
43
42
57
14
59
57
56
28
22
36
52
27
02
35
21
37
39
11
04
41
27
08
04
20
34
21
38
04
18
34
22
16
34
56
55
55
58
54
53
53
56
57
53
53
54
53
53
57
57
57
57
57
56
56
54
58
54
55
54
53
55
55
56
57
53
53
53
58
53
52
58
56
58
56
53
53
56
55
59
54
52
55
55
57
55
53
52
04
48
33
22
51
07
07
52
37
19
04
16
40
58
24
55
48
35
26
07
36
18
14
23
21
51
58
40
50
31
11
23
06
53
58
53
40
59
01
07
05
59
40
53
24
10
10
49
18
01
53
48
17
41
Badger
Bay d'Espoir Gen. Stn.
Bishop's Falls
Black Duck
Boat Harbour
Bonavista
Brownsdale
Buchans
Burgeo
Butlerville
Cape Race
Carmanville
Charleston
Clarenville
Cormack
Corner Brook
Cow Head
Daniel's Harbour
Deer Lake
Englee
Exploits Dam
Fogo
Gallants
Gander
Garnish
Glenwood
Goobies
Grand Falls
Harbour Breton
Harbour Deep
Hawke's Bay
Heart's Content
Holyrood Gen. Stn.
Indian Bay
Isles Aux Morts
Lethbridge
Logy Bay
Lourdes
Main Brook
McIvers
Middle Arm
Musgrave Harbour
North Harbour
Plum Point
Port Leamington
Port Aux Basques
Port Blandford
Portugal Cove
Rattling Brk. Norris Arm
Red Harbour
Rocky Harbour
Sagona Island
Salmonier Nature Park
Signal Hill
31
49
51
46
47
48
47
48
47
47
47
46
23
55
37
32
53
33
46
25
08
56
56
54
52
58
54
53
53
53
55
53
05
10
44
33
13
58
13
32
20
60.4
42.0
15.2
29.9
48.2
29.8
13.8
27.0
29.7
27.0
Sop's Arms White Bay
St. Anthony
St. Bride's
St. John's
Stephenville
Swift Current
Terra Nova Nat. Park
Victoria
Whitbourne
Winterland
FIGURE 2. Rainfall (mm) during Hurricane Michael (06 UTC 19 October - 06 UTC 21
October 2000). 3-hourly track positions shown by asterisks. Rain gauge sites shown by
circles.
32
FIGURE 3. Peak winds (km/h) during Hurricane Michael - 19 and 20 October 2000.
33
5. Coordination between the Hurricane Centres and the Newfoundland Weather Centre
In the case of the ship report of 80-knot winds at 1700 UTC 19 October, the forecaster at the
NHC phoned the duty forecaster at the CHC to discuss the credibility of the report. At the time he
had not yet confirmed the 80-knot wind but indicated that he would make another phone call after
consulting with the U.S. Coast Guard. The call was not received, so the forecaster at the CHC
transmitted the 1800 UTC bulletin, making note of the “unconfirmed” ship report but not upgrading
Michael’s winds. The CHC immediately relayed this information to the NWC in Gander, and later
issued an updated bulletin at 2100 UTC after calling the NHC to get final word on the ship report.
Just prior to the 0000 UTC 20 October update of the FXCN05, the CHC received another call from
the NHC regarding the time when Michael became extratropical. Both forecasters agreed that
Michael was extratropical not long after landfall, but they were unable to resolve the time of
transition. Frequent contact was made between the CHC and NWC just prior to and just after
landfall with regard to forecasts, warnings and details on Michael. The NWC contacted the CHC
with estimated time and location of landfall, and reported that the eyewall was detected on
Holyrood radar. Also, a draft of the WOCN10 CYQX was faxed to the CHC to confirm the data
before transmission. The CHC called NWC regarding the central pressure at landfall - estimated to
be 965 mb based on extrapolation from buoy 44255’s report of 967.7 mb. The bulletin was
finalized and transmitted at 9:49 p.m. local time.
6. Conclusions
Hurricane Michael presented many forecasting challenges during its rapid deepening and
extratropical transition over Newfoundland. Even during the early stages of its development from a
subtropical low, forecasters were unsure how intense the storm would become. The key lesson to
be learned from this event was the role of the upper-level mid-latitude trough in the deepening and
extratropical transition of the hurricane. Since the trough was intensifying (or “digging”) at
precisely the same time as Michael was moving just ahead of it, baroclinic energy was transferred
to the system in much the same way as it would be for an ordinary developing extratropical low.
The Canadian weather offices involved with the forecasting of this event (Canadian
Hurricane Centre, Newfoundland Weather Centre and Maritimes Weather Centre) issued many
warnings. In general the warnings were adequate, but forecast lead times were short owing to the
rapid forward acceleration of the storm. Forecasters are generally aware of this rapid motion, yet
Michael still caught most by surprise. Also surprising was the intensity of the storm at landfall,
partly due to the fact that there was little time for it to weaken over colder waters.
It goes without saying that there is much to be learned about forecasting extratropical
transition, particularly the potent combination of mid-latitude troughs with hurricanes. Detailed
studies of storms like this one, along with the analysis of aircraft reconnaissance and remotelysensed data integrated with numerical model simulations will continue to shape the future in our
understanding of these dangerous storms.
34
References
Richards, W. 2000: Atlantic Climate Centre, Fredericton, New Brunswick, Canada. [Contact
information available at http://www.ns.ec.gc.ca/climate/contacts.html]
Stewart, S. R., 2000: Tropical Cyclone Report Hurricane Michael 17 - 19 October 2000. 13 pp.
[Available online at http://www.nhc.noaa.gov/2000michael.html.]
35